Investigations of γ-secretase substrates Amyloid precursor protein and p75 neurotrophin receptor in zebrafish

Abstract

γ-secretase is an important protease complex responsible for the cleavage of over 100 substrates within their transmembrane domains. Among these substrates is the AMYLOID BETA (A4) PRECURSOR PROTEIN (APP), which is most well-known for its involvement in Alzheimer’s disease (AD) pathology and the p75 neurotrophin receptor (p75NTR). This thesis aimed to investigate γ-secretase and its substrates, p75NTR and APP, in a number of ways; Aim 1) to dissect the structural determinants contributing to selection of substrates by γ- secretase using an assay-based system, Aim 2) to investigate the molecular effects of fAD-like and null mutations in zebrafish appa and appb, by generating mutations in endogenous genes with genome editing technologies, and Aim 3) to investigate the aggregation propensity of Aβ-42-like peptides of Appa and Appb using predictive software. How γ-secretase cleavage site specificity is determined is still unclear. A previous study investigating the proteolytic processing of p75NTR and its homolog NRH1 (neurotrophin receptor homolog 1) found that transmembrane cleavage of NRH1 was not sensitive to the γ-secretase inhibitor DAPT, suggesting that it is not processed by γ-secretase. We have previously identified zebrafish orthologues of the p75NTR and Nrh1 genes and have developed in-vivo assays to assess cleavage of the p75NTR and Nrh1 proteins. To address Aim 1, we first improved upon our previous assay system by switching out the internal reference standard mCherry for a second GFP sequence. A chimeric construct was then designed, in which the Nrh1 transmembrane domain was replaced with the transmembrane domain of p75NTR, to allow us to determine whether this domain can confer γ-secretase cleavage susceptibility to Nrh1. Our results indicate that the p75NTR transmembrane domain alone is not sufficient to confer γ-secretase susceptibility to Nrh1. Missense mutations in the APP gene cause approximately 15% of dominantly inherited familial Alzheimer’s Disease (fAD). Currently 59 mutations are known and are listed on Alzforum (https://www.alzforum.org/mutations/app). The human APP gene has two co-orthologues in zebrafish, appa and appb. To address Aim 2, we attempted to utilise genome editing technology to generate fAD-like and null mutations in these zebrafish genes. Despite the various challenges associated with this project, a putative null mutation of the zebrafish appb gene was ultimately generated. Hypoxia is thought to be a risk factor for AD. A previous study in our laboratory measured the hypoxic response of 6-month-old zebrafish carrying fAD-like mutations in psen1 and observed an increased hypoxic response in these mutants under normoxia and a further increase under hypoxia. We measured the hypoxic response in our 6-month-old heterozygous appb putative null mutant zebrafish. However, there was no observable difference in the hypoxic response of our mutants compared to 6-month-old wild types. In future, our appb null mutation should serve to elucidate the role of this gene in neural function and its interactions with the other fAD genes. Aβ is proposed to be a key pathogenic molecule in AD. It is unknown whether APP genes in zebrafish form aggregation prone Aβ42-like peptides as occurs in the aging human brain. To address Aim 3 a bioinformatics approach was employed. Analyses using multiple software revealed that the predicted Aβ-42-like peptide of Appa zebrafish had similar aggregation propensity potential to that of human Aβ-42. The peptides we analysed consistently showed two domains of high aggregation propensity, one at the C-terminus and one in the middle of the peptide. Interestingly, the Aβ-42-like peptide of zebrafish Appb had comparable aggregation potential to human Aβ-42 in its C-terminal end but not in its mid-peptide region, suggesting that this peptide may not be as aggregation prone as human Aβ-42.